![[Game Analytics]](ga.png)
Proposal: Wednesday, April 24th, noon - Note time!
Report: Sunday, April 28th, 11:59pm
The goal of this project is to do some game analytics on a game of your choosing. You will plan out an analytics pipeline suitable for your selected game. This requires consideration of the game, the type and quantity of data that can be gathered from the game, needed playtesting (if any), and processing necessary to wrangle raw data into a form for analysis. In particular, you must plan ahead in terms of the analysis you seek, presented in the form of "hand drawn" charts, writing up your plan in the form of a written proposal. After having the proposal approved, you enact your plan to obtain and extract game data, analyze the data through charts and tables, and writeup the results in a report for dissemination.
This project can be done solo or in a group of two.
You must submit a proposal about 1/3 of the way in, leaving about 2/3 of the time until the project report is due. You will turn this in via Canvas (see Submissions below).
The purpose of the proposal is two fold: 1) to force you to "look before you leap" - investigating what a game of interest may provide for analytic opportunities before using it in a project; 2) allow for feedback on the appropriate scope and content of your project.
The proposal outlines the work that will constitute your project. The proposal should include:
A heading, indicating your project title (creativity welcome) and associated student name(s). Remember, this project can be done in a group of two, max.
A brief overview of the game used for your analytics. You can choose a brand new game or you can use one of the games from a previous projects (i.e., Heroes of the Storm, Mazetool, Hearthstone, or Pig variants).
The overview can be quite brief for one of the previously-used games, but should be more descriptive for any other game.
A brief description of the game analytics pipeline you will use, particularly if there are new scripts/tools needed to obtain the raw game data. You must include a brief description of the data available for analysis for the selected game. This should include details on the exact fields (e.g., game duration) as well as the native format (e.g., json).
A set of charts, hand drawn, that show your intended visual analysis. The charts should not have numbers nor necessarily even units, but should definitely have labeled axes and clearly depict trends/relationships you want to observe. For reference, see examples of hand drawn game analytics charts.
You must include at least 3 charts in the proposal and at least 3 different types of charts.
Tables can also be included, where appropriate.
For a bonus 2 points, test the normality of at least one data distribution with a normality probability plot (a quantile-quantile plot).
For a bonus 2 points, compare (with proper interpretation) the confidence intervals of two means.
Note, if you choose a previously analyzed game, you must do all new analysis.
A timeline for expected completion of major tasks. e.g., gather data, analyze game duration, analyze scores versus game type, ... write draft of report, write final report. The project time is only about a week (after the proposal), so the granularity should be day-by-day.
A grading guide of where points should be allocated for your project based on amount of effort it will require (e.g., game setup, data pipeline, analysis of players ...). The emphasis must prioritize the final analysis, but there is some leeway as to what parts of the analysis can be prioritized. See the grading guides for Project 1, Project 2, Project 3 and Project 4 for examples.
Note, your grade will be entirely reflected in the report, so for any effort awarded, it must be evident in the report.
Also note, you will submit a final version of your grading guide with your final report, revised to reflect the actual work and emphasis done.
The "size" of your project should be about the same as the previous projects, in terms of number of hours (i.e., plan on about 10-12 hours total, including proposal). But remember, unlike in previous projects, you need to "figure out" the analytics pipeline yourself! As such, this suggests slightly less analysis (i.e., fewer charts) than might have been produced in previous reports. Of course, the same does not hold for analysis if you use games from Projects 1, 2, 3 or 4 where the pipeline is already known. Similarly, if the game to be analyzed provides a .csv dataset (i.e., a simpler pipeline), more analysis should be done.
A project done in a group of two will need to be approximately 50% "bigger" than a project done individually.
You can analyze a game you have developed. However, the emphasis must be on analytics and not on instrumenting the game and definitely not on developing the game. Similarly, playtesting to gather data cannot be a major part of the project.
Your proposal text should be about 1-2 pages at most in length, but the entire document will likely be significantly longer with proposed charts, tables and any screen shots, as appropriate.
There are many possible games that are suitable. If in doubt, run ideas by the professor briefly (in person or via email) before starting work in earnest on the proposal.
A good way to find data of all types are:
Search terms (e.g., "games") can help narrow down choices.
Remember, when investigating a new game, account for extra time to figure out the data format (e.g., is it in json? are log files in some custom format?) and any needed scripts.
In the proposal, make sure to motivate the exploration for each type of analysis. Why are you doing this? What is the interesting question you are trying to answer? What is the hypothesis?
For the proposal (and final report, note that 3 charts of 3 different types is the minimum. More should be used for a deeper exploration, as appropriate.
For final reports, remember to describe the data sets (sampling, composition) and data pipeline (source, tools, filtering) in the report. Roughly, this length of this section in the report should be proportional to the effort in the project.
You writeup a report of your analysis.
The report should include a brief description of the game, as appropriate, and the methodology, particularly as it relates to the results obtained.
Make sure to have clearly organized report sections.
All guidelines for data analysis (charts, writeup, statistics) pertain to this report, too.
Remember to include a grading guide of where points should be allocated for your project. This should be based on the grading guide from your proposal, but can be adjusted somewhat based on where the actual emphasis of work/results ended.
You can re-use material from your proposal, as appropriate.
Both the proposal and the report are to be submitted electronically via Canvas by 11:59pm on the day due. There are separate entries for the Project 5 proposal and the Project 5 report.
A project done in a group should be submitted by only one member of the group, with the names of both group members clearly indicated on the proposal and report.
Documents must be a PDF, named proj5-lastname.pdf.
The submission is a report in PDF, named:
proj5-lastname.pdfwith your name in place of "lastname" above, of course.
To submit your assignment (proj5-lastname.pdf):
Open: IMGD2905-D19-D01
Navigate to:Assignments->Project 5
Click:Submit Assignment
Click:Choose File
Select the pdf file:proj5-lastname.pdf
Click:Submit Assignment
Important - you must click the Submit Assignment button at the end or your file will not be submitted!
When successfully submitted, you should see a message similar to:
Submission
- Submitted!
Apr 28 at 11:13pm
All accomplishments are shown through the documents (proposal and report). You will not submit scripts or data or any other aspects of your data analytics - just the proposal and report documents.
Proposal - 20%
Proposal document with details as described in this writeup.
Report - 80%
Final document, complete with writeup and analysis. Your report should be a full writeup, including revised grading guide.
100-90. The submission clearly exceeds requirements. All parts of the project have been thoroughly completed. The proposal and report are clearly organized and well-written, charts and tables are clearly labeled and described and messages provided about each part of the analysis.
89-80. The submission meets requirements. All parts of the project have been completed or nearly completed. The proposal and report are organized and well-written, charts and tables are clearly labeled and described and messages provided about each part of the analysis.
79-70. The submission barely meets requirements. Most parts of the project are completed or nearly completed, but some may be missing. The proposal and report are somewhat disorganized or not well-written. Charts and tables are clearly labeled, but there may be issues with the description and messages provided about each part of the analysis.
69-60. The project fails to meet requirements in some places. Not all parts of the project are completed. The proposal and report are disorganized and not well-written. Charts and tables are not clear nor up to standards described in class. Messages are not always provided for the analysis.
59-0. The project does not meet requirements. Many parts of the project are not completed. The proposal and report are disorganized and poorly written. Charts and tables are not clear nor up to standards described in class. Messages are generally not provided for the analysis.
The comments below are in response to reviewed proposals. The feedback is intended to provide general guidance for items that should be incorporated into the final report. They are not provided in any particular order.
Games chosen for analysis: League of Legends (x3), Pokemon (x2), Dota 2 (x2), PUBG (x2), Path of Exile, Fortnite, NBA, Overwatch
When referencing a game in a report, good practice is to include the publisher and the year published. e.g., League of Legends (Riot Games, 2009).
For final reports, remember to describe the data sets (when gathered, who gathered, what games, composition) and data pipeline (source, tools, filtering) in the report. Roughly, this length of this section in the report should be proportional to the effort in the project.
Make sure to motivate the exploration for each type of analysis. Why are you doing this? What is the interesting question you are trying to answer? What is to be potentially gained by the analysis? What is the hypothesis?
Note, 3 charts of 3 different types is the minimum. More should be used for a deeper exploration, as appropriate.
Return to the IMGD 2905 home page